Hi guys, please see topic for question.

 

Regards

Nokoff

 

I see the question, I understand what you are asking in the context of a normal RAID array, but the question makes no sense when applied to unraid. Unraid works much differently, and the only "waste" is space on the parity drive that is not reflected in a data drive.

 

1. Parity drive must always be equal to or bigger than the largest data drive.

 

2. All other drives are completely available for storage, so adding a drive adds the entire size to the available space.

 

So, if you already have a 4TB parity drive, adding a 4TB data drive incurs no loss, you get to use the entire drive.

 

If you don't already have a 4TB parity drive, you will have to use that 4TB as the parity drive, and you will gain the use of your current parity drive when you add it as a data drive (preferably after a preclear cycle, otherwise you will be down a long time while the drive is cleared. So you will "lose" the net difference between your old parity drive and the 4TB.

 

Clear as mud?

No.

 

The parity drive must be greater than or equal to the largest data disk in the array. The 4TB drive must be placed in the parity location and the parity drive can be formatted as a data disk. This will result in a gain of 2TB of storage. The next 4TB drive added to the array will result in a gain of 4TB of storage if none of the exiting disks are removed.

Just remember, your parity drive needs to be the largest drive.

So using the 4TB drive as your parity drive does not buy you any additional storage but does allow you to add other 4TB drives as data drives later. (if you are running unRaid 5.x)

Just remember, your parity drive needs to be the largest drive.

Actually, it has to be equal to or larger than any data drive.

No.

 

The parity drive must be greater than or equal to the largest data disk in the array. The 4TB drive must be placed in the parity location and the parity drive can be formatted as a data disk. This will result in a gain of 2TB of storage. The next 4TB drive added to the array will result in a gain of 4TB of storage if none of the exiting disks are removed.

 

I'm a little confused by what was said by the others who replied now. Will I or will I not waste 2TB of space on the first 4TB drive I put in my array? Can you indeed use the 2TB as storage?

 

Edit: Wait I think I understand where the confusion is, I don't HAVE a parity drive! at this moment. So the 2TB you are talking about is if I swapped out a 2TB Parity for a 4TB. So it is a wasted 2TB. I hate this! That's an additional cost of a thousand Rand, in addition to the cost of the unraid software, more money, less storage.

 

I'm going to see if I can sell my current drives in an effort to bring the LCD up to 3-4TB so I don't have to waste that space.

I'm a little confused by what was said by the others who replied now. Will I or will I not waste 2TB of space on the first 4TB drive I put in my array? Can you indeed use the 2TB as storage?

 

Not sure what's confusing, but I'll try to clarify for you.

 

The parity drive does NOT provide ANY storage for the array.  It's only used to store the parity information that enables UnRAID to be fault-tolerant.    Is that "wasted" ??  I'd say not, as it's providing a useful function (fault-tolerance).    But it's true that the parity drive provides ZERO actual storage for data.

 

Now, if you add a 4TB drive to an array that's currently all 2TB drives, you will have to use it as the parity drive, since the parity drive must be at least as large as all other drives.    After you do this, you can, of course, use the 2TB drive that WAS the parity drive as another data drive.  So you'll gain a net 2TB of data storage in that process.

 

Once you have  4TB parity drive, then you can add additional 4TB drives and all of the space will be available for your array.

 

There IS one situation where the above (and what everyone else said) is NOT true:  IF you have built an array WITHOUT parity then you can simply add the 4TB drive and you'll gain 4TB of additional storage.    But I assume that since you're running UnRAID, you DO have a parity drive -- an assumption which all of the answers (including my comments above) have also been based on 

Very interesting -- I wrote the last paragraph above just because I always try to be as complete as possible in my answers ... but I actually didn't think there was much chance that was actually the case 

 

But your edit (posted while I was writing it) shows that we should never make assumptions ... because you indeed DO have an array without a parity drive !!

 

As for "wasting space" ... you're not really doing that.  No matter what the size of your other drives, your parity drive will always be -- from the perspective you're looking at it -- "wasted space", since it provides no storage capacity.    But ALL of your other drives, of whatever size, contribute their full size to the array.    One of the nice advantages of UnRAID is that it allows you to do exactly that ... mix drives of varying sizes, so you don't have to buy all new drives every time you want to upgrade to the newest, higher-capacity drives.    Your parity drive is simply the upper limit on how big those drives can be.  So, for example, when 5TB drives come out, if you want to use those in your array, the first thing you'll need to do is get a 5TB parity drive.

 

I hear you but effectively if you use a drive larger than your 2nd biggest drive that is purely a waste, because you could accomplish the same thing with a smaller drive.

 

Basically it's optimum to have the same drive sizes all round.

I hear you but effectively if you use a drive larger than your 2nd biggest drive that is purely a waste, because you could accomplish the same thing with a smaller drive.

 

Basically it's optimum to have the same drive sizes all round.

 

Yes, it's indeed optimum to have all drives the same size.  However, most folks consider it an advantage to be able to mix sizes ... this allows you to use "extra" drives you have for additional storage space on your UnRAID array.  For example, you have 6 2TB drives.    With a traditional RAID-5 controller (UnRAID is not RAID-5, but in concept it's very similar), if you wanted to upgrade to larger drives, you'd have to buy SIX of them.    With UnRAID, you can buy ONE ... and then upgrade the other drives on an as-needed basis.

 

... and you wouldn't "accomplish the same thing" if you used a smaller drive.  For example, if you decided to add a parity drive, and used a 2TB parity drive, it would work => but you'd be limited to 2TB drives for all other drives you may want to add to the array.    If you use a 4TB parity drive, it will also work;  but will allow you to add drives up to 4TB in size (or replace your existing drives with 4TB drives one-a-a-time).

 

Don't think of the parity drive as "wasted space" ... think of it as "insurance".  THAT is what makes UnRAID fault-tolerant.  You're not (apparently) using UnRAID for fault-tolerance ... so in your case you're simply not "paying the price" for the fault-tolerance feature.

You may have already realized unRAID is different than from other traditional RAID arrays in that data (and parity) is not spread across several drives.

 

In your system if one of your drives fails you have no way to recover that data. Each file is essentially written to only one drive. Having a drive dedicated to parity gives you redundancy and a way to recover after a drive failure.

 

I hope that helps.

 

 

 

Thanks guys i think you've given me a bit of a more glass half full way of looking at it